A Landscape Restoration Plan for the South Willapa Bay Conservation Area

Widespread intensive forest management in the Willapa Hills of southwest Washington, USA, has altered landscape patterns, forest stand structure and ecological processes.  At present, less than 5% of the region remains in unmanaged or late-successional forest habitat. Because of the biological significance and rarity of old-growth Sitka spruce – western hemlock – western redcedar forests,  the Nature Conservancy and U.S. Fish and Wildlife Service have formed a partnership to restore landscape and stand level ecological processes on a 6800ha forested area. A landscape restoration plan was developed to achieve four broad desired future conditions: increased ecosystem resistance and resilience to disturbance at multiple scales, functional linkages between aquatic and terrestrial ecosystems, landscape heterogeneity consistent with that expected under a “natural” disturbance regime, and provision of habitat for late-successional associated species and species of concern. The restoration design employs two primary strategies: silvicultural treatments in previously harvested stands to promote the development late-successional characteristics and biological complexity, and the removal and improvement of the forest road network.  A landscape scale framework for prioritizing and coordinating stand and road treatments was developed.  The plan incorporates the long term restoration experiment being implemented in part of the project area. To measure progress towards the desired future conditions, treatments scheduled in the plan were modeled using the Landscape Management System (LMS) for 50 years. An index of late-successional forest structure and a wind risk model were customized and built into LMS. Results from the modeling suggest that silvicultural treatments will increase the number of large diameter trees, accelerate vertical canopy  development, not affect the number of large snags, and reduce overall dead wood accumulation. Overall wind resistance remains unchanged from thinning over the 50 year simulation period.